On Wednesday, Anthropic released a report detailing how Claude was misused during March. It revealed some surprising and novel trends in how threat actors and chatbot abuse are evolving and the increasing risks that generative AI poses, even with proper safety testing. In one case, Anthropic found that a "sophisticated actor" had used Claude to help scrape leaked credentials "associated with security cameras" to access the devices, the company noted in the announcement. In another case, an individual with "limited technical skills" could develop malware that normally required more expertise. Claude helped this individual take an open-source kit from doing just the basics to more advanced software functions, like facial recognition and the ability to scan the dark web.

Archetypal analysis (AA) was originally proposed in 1994 by Adele Cutler and Leo Breiman as a computational procedure to extract the distinct aspects called archetypes in observations with each observational record approximated as a mixture (i.e., convex combination) of these archetypes. AA thereby provides straightforward, interpretable, and explainable representations for feature extraction and dimensionality reduction, facilitating the understanding of the structure of high-dimensional data with wide applications throughout the sciences. However, AA also faces challenges, particularly as the associated optimization problem is non-convex. This survey provides researchers and data mining practitioners an overview of methodologies and opportunities that AA has to offer surveying the many applications of AA across disparate fields of science, as well as best practices for modeling data using AA and limitations. The survey concludes by explaining important future research directions concerning AA.

We review generalized additive models as a type of ``transparent'' model that has recently seen renewed interest in the deep learning community as neural additive models. We highlight multiple types of nonidentifiability in this model class and discuss challenges in interpretability, arguing for restraint when claiming ``interpretability'' or ``suitability for safety-critical applications'' of such models.

Existing distribution compression methods, like Kernel Herding (KH), were originally developed for unlabelled data. However, no existing approach directly compresses the conditional distribution of labelled data. To address this gap, we first introduce the Average Maximum Conditional Mean Discrepancy (AMCMD), a natural metric for comparing conditional distributions. We then derive a consistent estimator for the AMCMD and establish its rate of convergence. Next, we make a key observation: in the context of distribution compression, the cost of constructing a compressed set targeting the AMCMD can be reduced from $\mathcal{O}(n^3)$ to $\mathcal{O}(n)$. Building on this, we extend the idea of KH to develop Average Conditional Kernel Herding (ACKH), a linear-time greedy algorithm that constructs a compressed set targeting the AMCMD. To better understand the advantages of directly compressing the conditional distribution rather than doing so via the joint distribution, we introduce Joint Kernel Herding (JKH), a straightforward adaptation of KH designed to compress the joint distribution of labelled data. While herding methods provide a simple and interpretable selection process, they rely on a greedy heuristic. To explore alternative optimisation strategies, we propose Joint Kernel Inducing Points (JKIP) and Average Conditional Kernel Inducing Points (ACKIP), which jointly optimise the compressed set while maintaining linear complexity. Experiments show that directly preserving conditional distributions with ACKIP outperforms both joint distribution compression (via JKH and JKIP) and the greedy selection used in ACKH. Moreover, we see that JKIP consistently outperforms JKH.

The new Matter standard is getting better at helping Alexa, Apple HomeKit, Google Home, and Samsung SmartThings play nice with each other, but it often does so at the expense of finer-grained features. Some Matter-enabled smart lights, for example, can be turned on or off via Matter or change their color, but Matter controllers might not be able to access their lighting scenes or advanced animation modes. Likewise, smart home manufacturer Aqara found some of its hardware functionality hamstrung by Matter's limitations, such as the lack of Matter support for facial recognition (which might arrive once Matter finally works with security cameras), or for the fall-detection capabilities of its motion sensors. One option would be to wait for the Matter specification to catch up and add that functionality--which could take a while, given the slow pace of Matter specification updates. Instead, Aqara built its own workaround, which involves taking various Aqara scenes and "signals" and turning them into virtual sensors that Matter understands.

Power grid operation is becoming increasingly complex due to the rising integration of renewable energy sources and the need for more adaptive control strategies. Reinforcement Learning (RL) has emerged as a promising approach to power network control (PNC), offering the potential to enhance decision-making in dynamic and uncertain environments. The Learning To Run a Power Network (L2RPN) competitions have played a key role in accelerating research by providing standardized benchmarks and problem formulations, leading to rapid advancements in RL-based methods. This survey provides a comprehensive and structured overview of RL applications for power grid topology optimization, categorizing existing techniques, highlighting key design choices, and identifying gaps in current research. Additionally, we present a comparative numerical study evaluating the impact of commonly applied RL-based methods, offering insights into their practical effectiveness. By consolidating existing research and outlining open challenges, this survey aims to provide a foundation for future advancements in RL-driven power grid optimization.

Computational phenotyping is essential for biomedical research but often requires significant time and resources, especially since traditional methods typically involve extensive manual data review. While machine learning and natural language processing advancements have helped, further improvements are needed. Few studies have explored using Large Language Models (LLMs) for these tasks despite known advantages of LLMs for text-based tasks. T o facilitate further research in this area, we developed an evaluation framework, Evaluation of PHEnotyping for Observational Health Data (PHEONA), that outlines context-specific considerations. W e applied and demonstrated PHEONA on concept classification, a specific task within a broader phenotyping process for Acute Respiratory Failure (ARF) respiratory support therapies. From the sample concepts tested, we achieved high classification accuracy, suggesting the potential for LLM-based methods to improve computational phenotyping processes.

Given two boundary distributions, the Schr\"odinger Bridge (SB) problem seeks the ``most likely`` random evolution between them with respect to a reference process. It has revealed rich connections to recent machine learning methods for generative modeling and distribution matching. While these methods perform well in Euclidean domains, they are not directly applicable to topological domains such as graphs and simplicial complexes, which are crucial for data defined over network entities, such as node signals and edge flows. In this work, we propose the Topological Schr\"odinger Bridge problem (TSBP) for matching signal distributions on a topological domain. We set the reference process to follow some linear tractable topology-aware stochastic dynamics such as topological heat diffusion. For the case of Gaussian boundary distributions, we derive a closed-form topological SB (TSB) in terms of its time-marginal and stochastic differential. In the general case, leveraging the well-known result, we show that the optimal process follows the forward-backward topological dynamics governed by some unknowns. Building on these results, we develop TSB-based models for matching topological signals by parameterizing the unknowns in the optimal process as (topological) neural networks and learning them through likelihood training. We validate the theoretical results and demonstrate the practical applications of TSB-based models on both synthetic and real-world networks, emphasizing the role of topology. Additionally, we discuss the connections of TSB-based models to other emerging models, and outline future directions for topological signal matching.

The Model Context Protocol (MCP) is a standardized interface designed to enable seamless interaction between AI models and external tools and resources, breaking down data silos and facilitating interoperability across diverse systems. This paper provides a comprehensive overview of MCP, focusing on its core components, workflow, and the lifecycle of MCP servers, which consists of three key phases: creation, operation, and update. We analyze the security and privacy risks associated with each phase and propose strategies to mitigate potential threats. The paper also examines the current MCP landscape, including its adoption by industry leaders and various use cases, as well as the tools and platforms supporting its integration. We explore future directions for MCP, highlighting the challenges and opportunities that will influence its adoption and evolution within the broader AI ecosystem. Finally, we offer recommendations for MCP stakeholders to ensure its secure and sustainable development as the AI landscape continues to evolve.

While most modern machine learning methods offer speed and accuracy, few promise interpretability or explainability -- two key features necessary for highly sensitive industries, like medicine, finance, and engineering. Using eight datasets representative of one especially sensitive industry, nuclear power, this work compares a traditional feedforward neural network (FNN) to a Kolmogorov-Arnold Network (KAN). We consider not only model performance and accuracy, but also interpretability through model architecture and explainability through a post-hoc SHAP analysis. In terms of accuracy, we find KANs and FNNs comparable across all datasets, when output dimensionality is limited. KANs, which transform into symbolic equations after training, yield perfectly interpretable models while FNNs remain black-boxes. Finally, using the post-hoc explainability results from Kernel SHAP, we find that KANs learn real, physical relations from experimental data, while FNNs simply produce statistically accurate results. Overall, this analysis finds KANs a promising alternative to traditional machine learning methods, particularly in applications requiring both accuracy and comprehensibility.